Adversarial Training with Synthetic Misspellings for Cross-Lingual Dense Retriever Generalization on LoCoMo

Effective cross-lingual dense retrieval methods that rely on multilingual pre-trained language models (PLMs) need to be trained to encompass both the relevance matching task and the cross-language alignment task. However, cross-lingual data for training is often scarcely available. In this paper, rather than using more cross-lingual data for training, we propose to use cross-lingual query generation to augment passage representations with queries in languages other than the original passage language. These augmented representations are used at inference time so that the representation can enco

Research goal: How does the incorporation of adversarial training with synthetic misspellings impact the generalization performance of dense retrievers across different languages in the LoCoMo benchmark?

Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 8.5/10.